Will A Robot Steal Your Job?

Submitted by Erico Matias Tavares of Sinclair & Co

Some Thoughts on Automation

Will a robot steal your job?

It turns out that the answer depends on the prevailing macroeconomic conditions much more than people think.

We started our professional career at General Electric, when it was still being run by the legendary Jack Welch. At that point GE’s experience with automation had not been fantastic. It had been tried in several manufacturing facilities but managers invariably found that rather than improving the bottom line the new robots were creating increasing amounts of waste. In other words, they were making the same mistakes as before – only faster.

The focus then shifted to analyzing key business processes as a whole to figure out where they could be improved. In 1995 Welch implemented an ambitious Six Sigma program with the goal of massively reducing defects across all of GE’s businesses (a six sigma process only has 3.4 errors per million opportunities). In fact he made it central to the business strategy of the company, tying executive compensation to Six Sigma results and making sure that everyone at all levels got adequately trained on the concept.

As employees across the organization began figuring out opportunities for improvement, automation became a tool rather than the focus of the entire exercise.

In the finance department of a division we worked in, we found that building a simple Microsoft Excel macro to extract data from different data sources massively reduced opportunities for error, allowing us to spend more time on value added projects. Not advanced robotics stuff, but effective nevertheless. That extra time for us was a huge benefit of automation, making work a lot more stimulating and rewarding.

In another GE division we worked in, there was a gentleman by the name of Bob who probably had one of the most tedious jobs on the planet. Every working day from 7am to 3pm he would be continuously handed baskets of bolts, which he would then proceed to sort manually with the assistance of a monocle (earning him the nickname of “Bob the Bolt”). To make things even more tedious, there were only three types of bolts, all with a similar size (something to consider when you think your job sucks).

Bob was incredibly proficient at his job. We estimated that he could sort up to 80 bolts per minute. He did not seem to mind the task either, quite the contrary. And at 3pm he was out the door while most of us were laboring away on those value added projects.

So here’s an interesting thing about automation: for us it was a blessing, allowing us to do more with our time and avoid truly menial tasks; for Bob it might have been a disaster, as he could lose his job (although we could certainly argue that his skills would be much better employed elsewhere in the shop floor). Whether it is a good or a bad thing largely depends on individual preferences and employment circumstances.

This question became much more intricate as machines evolved over time, giving them a much broader range of capabilities at a higher affordability.

Decades ago Henry Ford devised new methods that greatly reduced production costs, making his cars affordable to the vast majority of Americans at long last. His genius was figuring out that specialization, where a worker would be responsible for a single task along a production line, rather than producing certain components of that line per se, greatly increased productivity. This in turn massively reduced costs per unit.

The problem, however, is that work became incredibly boring and repetitive. As a result, morale dropped and employee turnover exploded, to the point where hiring and training new employees became very costly because they seldom stuck around long enough for the company to get a payback on that investment.

Ford eventually figured out a solution for this problem too: substantially raise wages. Not only were employees more motivated as a result, they now had additional purchasing power to buy his cars as well. The company’s efforts to increase productivity increased the size of the overall pie, enabling workers to receive a better pay.That productivity increase arguably would have been even greater if industrial technologies had been available to assist employees back then, providing much better working conditions (which unfortunately remained very challenging during the first half of the 20th century).

Today we have very capable machines at our disposal. A manager’s decision to automate the company’s production line rather than hire more people largely depends on the relative cost of both, adjusted for productivity. A productive team of employees is a formidable competitor to any robot, with lateral problem solving capabilities and creativity to boot. Bob had a job because he was so good at it that management decided that the cost of automating it was not justifiable.

As such, workers do not need to live in constant fear that their job will be eventually replaced by robots. In many instances their careers can become more productive and rewarding as a result of technology. Just ask any farmer in the US, where grain harvesting equipment has become a marvel of sophistication. Moreover, while computing power has been rising exponentially for decades, a robot’s ability to do manual tasks remains limited. For one, as far as we know making a robot “see” remains an unsolvable problem.

Unfortunately, not everything is rosy for workers. One thing they should be greatly concerned about is the law of unintended consequences – especially resulting from the policies implemented by the very people claiming to be helping them.

There is ample evidence of this in both developed and emerging markets:

  • In the US, by cheapening credit to levels never seen before the Federal Reserve has made capital highly affordable, skewing the economics against hiring people. Exploding healthcare and other regulatory costs can only make things worse. No wonder this recovery has been one of the worst ever in terms of job creation.
  • Workers in China are suffering from an even more aggressive monetary policy, which inflates their cost of living (thus requiring higher wages just to get by) and cheapens the relative cost of capital. Chinese factories employing robots (in some cases exclusively) regularly make headlines these days. The resulting gains from productivity improvements thus become less available to workers as a whole.

In other words, in a technologically-advanced economy, cheapening capital by central bank decree can actually destroy jobs rather than create them. Try wrapping your head around that one Thomas Piketty.

It seems that many of our jobs are under threat as a result of policies that create severe economic distortions, much more than the robots per se.

Bob the Bolt would surely agree.